CORE 1: COMPUTATIONAL SCIENCE Although there Is a strong methodological component to core 1, the substrate for the work contemplated here - large-scale health systems with all their quirks, biases, and heterogeneity of practice - Is by Its nature resistant to standard analytical approaches. However possible, novel methodologies for interpreting data derived from these noisy systems must provide derived data that can be documented as useful to clinical researchers, as represented by the principal investigators of our